Last Updated: September 22,2026 

 Editorial note: This article contains no sponsored placements. This article currently contains no affiliate links, and no vendor mentioned here paid for or previewed it. If that changes, I will say so here and mark any affiliate links clearly, per our Affiliate Disclosure. Pricing was pulled from each vendor's public pricing page and cross-checked against third-party procurement data (Vendr, Tropic) between July and September 2026. Every dollar figure below has a source. Prices change — always confirm current numbers on the vendor's own page before you buy, and check the "verified" date next to each figure. Two of the figures in this article reflect limited-time promotions rather than standing list prices; both are labeled as promotional where they appear. I re-check this article's pricing quarterly and update it when a figure moves — see the "last verified" date above for the most recent pass.

Quick answer

Most SaaS tools now price on one of three models: per-seat (you pay per named user, like Slack), usage-based (you pay per unit consumed, like Zapier's per-task billing), or hybrid (you pay per seat and per usage meter, like HubSpot's seats-plus-contacts structure). None of the three is inherently cheaper. Each one has a specific, well-documented way of quietly inflating your bill — inactive seats on per-seat plans, workflow-step multipliers on usage plans, and stacked onboarding fees plus contact-tier creep on hybrid plans. Below, we price all three using live 2026 figures, show the exact mechanism behind the overpay pattern in each one, and give you a five-minute audit to check your own stack.

Why this matters more in 2026 than it used to

SaaS procurement platform Zylo, which tracks license usage across enterprise software portfolios, found that organizations leave 46% of purchased software licenses unused on average, and that this idle spend adds up to roughly $19.8 million a year in wasted SaaS spend across the portfolios it analyzed. That single number is the reason this article exists: pricing model determines exactly how that waste gets billed to you, and the three models hide it in three completely different places.

I am not going to give you generic advice about "auditing your software stack." We're going to open the actual pricing pages of three tools that represent each model — Slack (per-seat), Zapier (usage-based), and HubSpot (hybrid) — pull the live numbers, and show you the arithmetic.

The three models, defined plainly

Case study 1: Per-seat pricing — Slack

Slack is the textbook per-seat model: the bill is a flat rate multiplied by however many people have an account in your workspace, active or not.

Live pricing, verified September 2026 (source: Slack's official pricing page, cross-checked against Vendr's transaction database and Tropic's procurement data):

Slack's pricing page, captured live on September 22, 2026, 10:45 PM.

A live-pricing correction, caught by our own screenshot: the table above uses Slack's standard annual-billing list price ($7.25 Pro / $15 Business+). Our own screenshot, taken the same week this article was verified, shows something the annual-list-price table doesn't: Slack is currently running a 50%-off-for-3-months promotion on monthly billing, dropping Pro to $4.38/user/month and Business+ to $9/user/month for new signups — both shown struck through against the regular monthly rates of $8.75 and $18. That promotional rate reverts to the full monthly price after three months and is a different number from the discounted annual-commitment rate. This is exactly the gap this article is about: the number a buyer sees at signup and the number on the invoice six months later are frequently two different prices, and which one applies depends on a billing-cadence toggle most people don't notice mid-checkout.

The overpay pattern, with real numbers: Business+ pricing rose from $12.50 to $15.00 per user per month — a 20% increase — applied automatically at each customer's first renewal after August 17, 2025, according to pricing-tracking site Rock. The same price change folded Slack's previously separate AI add-on into the base plan, which means existing customers now pay for AI capability whether they use it or not.

Layer that onto the seat problem: Slack bills every account in the workspace, not every active account. A 100-person company where only 30 people open Slack in a given month is still billed for 100 seats unless someone manually runs a deactivation pass — Slack does offer a "fair billing" credit for inactive users, but only on annual contracts, and only if you request it.

Run the math on a mid-size team: a 250-person company on Business+ at $15/seat/month pays $45,000/year at full headcount. If Zylo's 46%-unused figure holds even loosely for that org, roughly $20,700 of that annual bill is paying for logins nobody makes.

What real users say about it. On G2, a construction-industry director reviewing a comparable HubSpot seat-based product wrote that the pricing is somewhat expensive, though the paid tier's added capability can justify the cost for some buyers. It's a pattern that shows up across every seat-billed SaaS category, not just Slack: reviewers rarely say the product is bad — they say the bill doesn't track who's actually using it.

Case study 2: Usage-based pricing — Zapier

Zapier prices by the task, not the seat: one task equals one successful action step inside an automated workflow ("Zap"). Triggers and filters are free; every action after that consumes one unit of your monthly allowance.

Live pricing, verified September 2026 (source: Zapier's pricing page, cross-checked against three independent 2026 pricing breakdowns):

Zapier's pricing page, captured live on September 22, 2026, 10:49 PM, pay-monthly view. Confirms Professional starting at $29.99/month and Team at $103.50/month — matching the figures in the table above.

The overpay pattern, with real numbers. The mechanism is the task multiplier, and it's the single most under-explained thing in usage-based SaaS pricing: a Zap that looks simple to a buyer — "when a new lead comes in, notify the team" — is rarely one task in practice. A standard lead-routing workflow that enriches a lead, writes it to a CRM, posts to Slack, and sends a confirmation email is four tasks per run, not one. At 30 leads a day, that's roughly 3,600 tasks a month from a single automation — more than a third of the entire 2,000-task Professional tier consumed by one workflow.

Scaled up, the effect compounds dramatically. Independent pricing analysis puts Zapier's cost at roughly $72,000/year at 2 million tasks/month — a volume that sounds enterprise-scale until you remember that a handful of active multi-step Zaps running continuously can reach six figures of monthly tasks faster than most teams budget for. That's precisely the volume point at which teams typically start evaluating Make or self-hosted n8n as lower per-task alternatives, according to multiple 2026 automation-cost comparisons.

The buyer-side failure here isn't Zapier hiding the task-count math — it publishes it. It's that teams price their automation plan off the simple version of their workflow during the sales conversation, then build the real, multi-step version after they've signed, and don't revisit the plan tier until the overage charges show up.

Case study 3: Hybrid pricing — HubSpot

HubSpot is the clearest hybrid example on the market: you pay per seat, per marketing contact (a second, independent meter), and — critically — a mandatory one-time onboarding fee that most buyers don't see until checkout.

Live pricing, Marketing Hub, verified September 2026 (source: HubSpot's official pricing page, cross-checked against Agiled's and Zeeg's independent 2026 pricing breakdowns):

Additional seats on Professional run $45/month each; additional marketing contacts run roughly $50/month per extra 1,000 on Professional and $100/month per extra 10,000 on Enterprise.

HubSpot's Marketing Hub pricing page, captured live on September 22, 2026, 10:51 PM.

A live-pricing note from the same screenshot: at the moment of capture, HubSpot was running a "Save up to 65% on Starter" promotion, showing Starter at $10/mo/seat on annual billing, struck through against the regular $20/mo/seat rate — consistent with the $15–$20 range in the table above, just at the deep end of a limited-time offer rather than the standing list price. The screenshot also confirms Starter's included allowance directly from the vendor: 1,000 marketing contacts and 500 HubSpot credits, matching what's reported in the table. HubSpot marks this offer "new customers only" and "available for a limited time," so confirm it's still live before quoting it to readers.

The overpay pattern, with real numbers. Two things compound here, and both are documented directly on HubSpot's own pricing structure:

1. The tier cliff. The jump from Starter ($20/seat/month) to Marketing Hub Professional ($890/month base) is roughly a 44x increase in list price, according to pricing analysis site Tinycommand — and it's not gradual. There's no $100 or $300 middle tier; a growing team either stays on Starter's limited automation or absorbs the full Professional jump plus the one-time $3,000 onboarding charge in month one.
2. The dual meter. Seats and contacts scale independently. A seven-person marketing team on Professional adds $180/month in extra seats on top of the base $890 — and if that same team's contact list grows past the 2,000-contact allowance (which happens fast for any active email program), the contact-tier overage stacks on top of that. HubSpot's own contact-tier table shows the monthly Professional price climbing from $890 at 2,000 contacts to roughly $4,100 at 100,000 contacts — before a single additional seat is added.

What real users say about it. A verified HubSpot user reviewing on G2 (reprinted via AWS Marketplace) wrote plainly that the price is quite expensive, making it hard for small companies to use its best features fully. Another Sales Hub reviewer on G2 put it more bluntly, describing HubSpot's cost structure as one where the high price limits use for new businesses that cannot afford the investment. A third-party review aggregator that compiled hundreds of HubSpot reviews found that "High Costs and Lack of Transparency" was one of the most consistently repeated complaint categories, with reviewers specifically flagging that key functionality requires purchasing additional packages on top of an already premium-priced subscription.

That last point is the hybrid model's signature overpay mechanism: it isn't that any single number on the pricing page is dishonest. It's that three independently-scaling charges — seats, contacts, and a mandatory onboarding fee — are disclosed in three different places, and almost nobody adds them up before signing.

Why vendors design tiers this way (it's not an accident)

It's worth understanding the incentive on the other side of the pricing page, because it explains why the overpay pattern in each model is so consistent across the industry rather than a quirk of these three specific vendors.

Per-seat pricing rewards the vendor for slow deprovisioning. Every day a departed employee's account stays active is a day of billed revenue with zero marginal cost to the vendor — the software doesn't get slower or more expensive to run because an unused seat exists. That's precisely why "fair billing" credits, where they exist at all, are opt-in rather than automatic: automatic reclamation would cut directly into recurring revenue that costs the vendor nothing to keep collecting.

Usage-based pricing rewards the vendor for underestimated unit counts at the sales stage. A sales demo almost always shows the simplest possible workflow — one trigger, one action — because that's the version that's easy to explain and cheap-looking on the pricing calculator. The real, multi-step version that a team actually needs in production is assembled after the contract is signed, during implementation, by which point switching plans (or switching vendors) carries real friction. The gap between the demo workflow and the production workflow is not a bug in the sales process; it's structurally where the upsell lives.

Hybrid pricing rewards the vendor for buyers who evaluate only one meter. A buyer comparing HubSpot to a competitor typically anchors on the headline monthly price and the seat count, because those are the numbers that show up first in a sales conversation and in most comparison articles. The second meter — contacts, records, API calls — grows quietly with the customer's own business success, which means the price increase feels less like a vendor decision and more like a natural consequence of growth. That framing makes the increase far less likely to trigger a renegotiation than a flat price hike would.

None of this makes per-seat, usage-based, or hybrid pricing dishonest as a category. Every mechanism described above is disclosed somewhere on the vendor's own pricing page or terms of service — the Slack seat-increase notice, Zapier's published task-per-action documentation, HubSpot's contact-tier table. The overpay isn't concealment; it's asymmetry of attention. The vendor's pricing team has modeled exactly how the meter climbs. Most buying teams model it once, at signup, and then never again until the renewal invoice is already higher.

The overpay pattern, summarized

Putting the three case studies side by side:

The common thread across all three: the headline number on the pricing page is real, but it is never the number on your actual invoice. Per-seat hides waste in headcount. Usage-based hides it in unit multipliers. Hybrid hides it by spreading the true cost across meters that are disclosed on different pages.

A five-minute audit for your own stack

1. For every per-seat tool you own: pull the admin panel's last-login report. Anything inactive 30+ days is a candidate for downgrade or reclamation before your next renewal — not after.
2. For every usage-based tool you own: find your actual average units-per-workflow, not the "one trigger, one action" example from the sales demo. Multiply by real daily volume, not a pilot-period estimate.
3. For every hybrid tool you own: list every meter separately (seats, records, contacts, API calls) and check whether each one is trending toward its next tier threshold independently. Vendors size tiers to be crossed, not to be comfortable margins.
4. Before any renewal: ask the vendor directly whether a "fair billing" or inactive-seat credit exists — most per-seat vendors have one, and most sales reps won't mention it unless asked.

Which model actually suits your team?

• Choose per-seat if your headcount is stable and most people who have access actually use the tool daily — the classic case is a small, dense team where usage tracks headcount closely.
• Choose usage-based if your workload is spiky or automatable and you can forecast volume with some confidence — but budget for the real per-workflow unit count, not the simplest-case example.
• Expect hybrid if you're buying a platform (CRM, marketing automation, support desk) rather than a tool — and budget the onboarding fee and the second meter into year-one cost from the start, not as a surprise at checkout.

Frequently asked questions

Is per-seat pricing always more expensive than usage-based? Not inherently — it depends entirely on your usage intensity relative to seat count. A small, highly active team is usually cheaper on per-seat; a large team with light, sporadic use is usually cheaper on usage-based, assuming the usage stays predictable.

Why do so many SaaS companies move to hybrid pricing? Because it captures value from both dimensions of growth — headcount and usage — without forcing the vendor to choose. From a buyer's side, that's exactly why hybrid bills are the hardest to forecast: two variables are moving at once.

What's the single most effective way to avoid overpaying? Run the five-minute audit above before every renewal, not after a price increase notice arrives. Every case study in this article shows a documented price change (Slack's Business+ increase, HubSpot's contact-tier steps, Zapier's overage rate) that was disclosed publicly well in advance — the overpay happens when nobody checks the notice against actual usage.

Sources and verification

• Original screenshots captured by the author on September 22, 2026: slack.com/pricing, zapier.com/pricing, hubspot.com/pricing/marketing
• Slack official pricing page, slack.com/pricing (accessed September 2026)
• Vendr, "Slack Cost Per User: Real Pricing by Plan in 2026" — 535 verified purchase records
• Rock, "Slack Pricing 2026: Plans, Real Costs and Free Limits"
• Tropic, "How much does Slack cost?" procurement glossary
• Zapier official pricing page, zapier.com/pricing (accessed September 2026)
• Toolradar, "Zapier Pricing in 2026: Every Tier, the Real Per-Task Cost"
• Tinycommand, "Zapier Pricing Explained (2026): Tasks, Tiers, and the Real Math"
• HubSpot official pricing page, hubspot.com/pricing (accessed September 2026)
• Agiled, "HubSpot Pricing 2026: Every Hub, Every Tier, and the Hidden Costs"
• Tinycommand, "HubSpot Pricing Explained (2026): The $20 Seat, the $890 Cliff"
• Zylo, "Reduce License Waste" — SaaS license utilization data
• G2.com verified user reviews (HubSpot Sales Hub, HubSpot Marketing Hub pricing pages)
• AWS Marketplace, reprinted G2 verified reviews for HubSpot

All prices were current as of the "last verified" date above. SaaS pricing changes frequently — always confirm live figures on the vendor's own pricing page before making a purchasing decision.

About the author: 

Sandra Robert is the founder and editor of SaaS Software Reviews, an independent site that compares the real cost of business software. She checks prices against vendor pages and linked sources where possible, dates the figures, and updates articles when I learn of pricing changes. Corrections? Contact Sandra.

Editorial standards: saassoftwarereviews.com does not accept payment from vendors in exchange for favorable coverage, and pricing figures in our articles are drawn from vendors' own published pages plus independent, named procurement sources — never estimated or invented. When a figure changes after publication, I update the article rather than leaving stale numbers live.

Affiliate disclosure: This article currently contains no affiliate links. See my Affiliate Disclosure for how this site may be funded in future.